Suhail(@Suhail)

There's nothing that I hate more these days than this moment.

Suhail (@Suhail) — Fri, 11 Sep 2026 03:30:01 GMT

There's nothing that I hate more these days than this moment.

💬5🔄2❤️21👀2431📊7

⚡ Powered by xgo.ing

Out of curiosity, how does one build something that coordinates 10K AI agents to solve a single prob...

Suhail (@Suhail) — Tue, 08 Sep 2026 19:05:48 GMT

Out of curiosity, how does one build something that coordinates 10K AI agents to solve a single problem?

💬329🔄39❤️1638👀293016📊487

⚡ Powered by xgo.ing

--- name: cleanbranch description: Create a new branch with only the essential feature changes. Use ...

Suhail (@Suhail) — Sun, 06 Sep 2026 17:32:47 GMT

---
name: cleanbranch
description: Create a new branch with only the essential feature changes. Use for /cleanbranch, $cleanbranch, or requests to extract a minimal feature branch.
---

# Clean branch

Create the smallest complete PR for the intended feature, with clear code and minimal comments.

- Infer the feature from the conversation and review existing changes, including committed,
  uncommitted, and untracked work.
- Create a fresh branch from the user's chosen base, the current PR's base, or the repository's
  default branch. Use a descriptive branch name unless one is provided. Prefer a separate worktree
  to preserve the original branch and working files.
- Carry over only what the feature needs to work end to end, including necessary integration
  and tests. Leave out unrelated refactors, formatting changes, experiments, and speculative work.
- Favor clear names, simple control flow, and existing patterns. Avoid unnecessary abstractions.
  Comment only on non-obvious intent or constraints; remove comments that repeat the code.
- Re-verify that the feature works end to end, maintains comparable performance and quality,
  and satisfies the original requirements and intended scope.
- Review the full diff against the base and run relevant tests and repository-required checks.
  Fix failures introduced by the extraction and report any validation limitations.
- Report the branch, base, worktree path, what was kept or omitted, and validation results.
  Commit, push, or open a PR only when requested or already authorized.

💬2🔄1❤️32👀8978📊15

⚡ Powered by xgo.ing

How I go from 100K LOC to a few thousand these days: Lately I find that I enjoy having a very messy...

Suhail (@Suhail) — Sun, 06 Sep 2026 17:31:14 GMT

How I go from 100K LOC to a few thousand these days:

Lately I find that I enjoy having a very messy branch with lots of twists and turns, tests for verification, scripts to load test/adversarially break things, and experiments. Sometimes I am trying things to understand the pieces better. Rarely do I want all that bloat and slop to enter the main branch of the repo.

Please copy or tell me how to do it better.

💬20🔄6❤️305👀37430📊123

⚡ Powered by xgo.ing

@martin_casado Nah, not a skill issue. I always try one-shotting things with /goal now to see if it ...

Suhail (@Suhail) — Sun, 06 Sep 2026 17:20:13 GMT

@martin_casado Nah, not a skill issue. I always try one-shotting things with /goal now to see if it jumps in capability and I'd say I was fairly disappointed.

💬4🔄2❤️106👀26319📊16

⚡ Powered by xgo.ing

I hope it is clear that copying and following the herd for too long will unlikely lead to greatness ...

Suhail (@Suhail) — Sun, 06 Sep 2026 14:56:40 GMT

I hope it is clear that copying and following the herd for too long will unlikely lead to greatness but it’s a good way to get started while avoiding imagined markets with no real customers.

💬4🔄1❤️42👀4334📊9

⚡ Powered by xgo.ing

I would encourage more founders to copy the ideas out there that you’re passionate about. Not becaus...

Suhail (@Suhail) — Sun, 06 Sep 2026 14:49:21 GMT

I would encourage more founders to copy the ideas out there that you’re passionate about. Not because that is your end game, but to the degree it is something you enjoy building, you can get some customers and cockroach. That will give you time to adapt and figure out what unique thing it can evolve to as technology progresses.

In the age of AI, people can build very quickly. So, in many situations, it’s best to be in market with a team, funding, and some customers to take advantage of new capabilities in the models as they come out. It’s much harder to do what we used to do: quietly build a thing for 3-6 months and launch it. By the time you get the company started, you’ll have moved too slowly.

You should assume you will look no different than everyone adjacent to you in the first 3-4 months. However it may only be  like that for 3-4 months.

💬43🔄26❤️570👀41645📊145

⚡ Powered by xgo.ing

🎯 There are very few original startup ideas. It’s mostly adapting and learning from the market as ...

Suhail (@Suhail) — Sun, 06 Sep 2026 00:51:06 GMT

🎯

There are very few original startup ideas. It’s mostly adapting and learning from the market as you go. Most often you need to prepare for the moment especially these days.

Paul Graham@paulg

I talked to a founder who was demoralized because all the ideas he could think of could be easily duplicated. I told him to ignore that worry; nearly all startup ideas are like that at first. Their value is the other ideas they lead to.

🔗 View Quoted Tweet

💬31🔄33❤️680👀93761📊106

⚡ Powered by xgo.ing

Man, AI is insane.

Suhail (@Suhail) — Thu, 03 Sep 2026 19:51:35 GMT

Man, AI is insane.

💬39🔄26❤️590👀25088📊89

⚡ Powered by xgo.ing

Excited to see benchmarks headed in this direction and getting better!

Suhail (@Suhail) — Wed, 02 Sep 2026 19:58:31 GMT

Excited to see benchmarks headed in this direction and getting better!

Proximal@ProximalHQ

We are releasing FrontierSWE v2, our updated ultra-long horizon coding benchmark

V2 features an expanded task suite and improved methodology. We see large performance gaps between frontier models, with Claude Fable 5.1 leading by a wide margin

🔗 View Quoted Tweet

💬2🔄1❤️42👀15724📊7

⚡ Powered by xgo.ing

why is everything AI writes 10 billion lines of code?

Suhail (@Suhail) — Wed, 02 Sep 2026 03:35:11 GMT

why is everything AI writes 10 billion lines of code?

💬131🔄6❤️318👀44998📊138

⚡ Powered by xgo.ing

Must be in SF. We work in-person. No exceptions for this role.

Suhail (@Suhail) — Tue, 01 Sep 2026 16:07:51 GMT

Must be in SF. We work in-person. No exceptions for this role.

💬3🔄0❤️5👀3159📊3

⚡ Powered by xgo.ing

Looking for an excellent product-facing AI engineer. You will work on the frontier of everything goi...

Suhail (@Suhail) — Tue, 01 Sep 2026 16:05:09 GMT

Looking for an excellent product-facing AI engineer. You will work on the frontier of everything going on in AI/LLMs and work directly with me. 

Must have interesting side projects you did out of joy.

If you know someone excellent, will give you $10K if we hire them. DM me.

💬31🔄15❤️232👀44956📊72

⚡ Powered by xgo.ing

Updated to iOS 27 beta and Siri finally uses an AI that is unlike living in 2018. It's not great but...

Suhail (@Suhail) — Sun, 30 Aug 2026 18:04:23 GMT

Updated to iOS 27 beta and Siri finally uses an AI that is unlike living in 2018. It's not great but it is truly better than whatever it was.

💬13🔄0❤️65👀24846📊17

⚡ Powered by xgo.ing

If you have an app that has 1K DAU, I will give you GLM 5.3 Flash for free up to 1B tokens a month. ...

Suhail (@Suhail) — Sat, 29 Aug 2026 18:52:18 GMT

If you have an app that has 1K DAU, I will give you GLM 5.3 Flash for free up to 1B tokens a month.   

This is a one-time thing for someone. We need a little bit of data on understanding something.   

DM. Tweet will be deleted.

💬5🔄1❤️21👀2940📊7

⚡ Powered by xgo.ing

A late night thought: there will be a scarce amount of hackers who know things quite deeply enough s...

Suhail (@Suhail) — Fri, 28 Aug 2026 07:24:22 GMT

A late night thought: there will be a scarce amount of hackers who know things quite deeply enough still that they can steer a strong AI to an even stronger outcome because they know what to ask, what to guide it toward, and how to end up solving the right thing.

💬43🔄11❤️212👀23708📊63

⚡ Powered by xgo.ing

Need 256 B200s/B300s stat. Cannot wait till next month. DM.

Suhail (@Suhail) — Thu, 27 Aug 2026 03:19:41 GMT

Need 256 B200s/B300s stat. Cannot wait till next month. DM.

💬44🔄6❤️169👀55584📊57

⚡ Powered by xgo.ing

GLM 5.3 Flash benchmarks 🫡 If you want US based inference for this, DM me. We are happy to suppo...

Suhail (@Suhail) — Wed, 26 Aug 2026 14:35:33 GMT

GLM 5.3 Flash benchmarks 🫡

If you want US based inference for this, DM me. 

We are happy to support US enterprises.

💬0🔄0❤️8👀798📊1

⚡ Powered by xgo.ing

If you want direct access to 0x model by a US provider, dm me. Tweet will be deleted.

Suhail (@Suhail) — Wed, 26 Aug 2026 13:37:55 GMT

If you want direct access to 0x model by a US provider, dm me.

Tweet will be deleted.

💬0🔄0❤️5👀523📊1

⚡ Powered by xgo.ing

17/ The software around the harness is the new browser. We had realized this week that in order to...

Suhail (@Suhail) — Tue, 25 Aug 2026 14:36:25 GMT

17/ The software around the harness is the new browser. 

We had realized this week that in order to support customers, we needed to test a bunch of these to make sure the api works kinda like testing IE, Firefox, etc

💬0🔄0❤️5👀1476📊1

⚡ Powered by xgo.ing

If you care about enterprise revenue, this is the thing all the engineers are going to be using. I ...

Suhail (@Suhail) — Tue, 25 Aug 2026 13:22:15 GMT

If you care about enterprise revenue, this is the thing all the engineers are going to be using.

I cannot imagine not managing agents from my phone vs needing my laptop around me all the time.

💬7🔄3❤️59👀7699📊12

⚡ Powered by xgo.ing

The entire Remote tab inside ChatGPT is its own product, company category, and business. To bury it ...

Suhail (@Suhail) — Tue, 25 Aug 2026 12:40:08 GMT

The entire Remote tab inside ChatGPT is its own product, company category, and business. To bury it inside a chat ai app is a blunder. It’s also very buggy!

💬70🔄22❤️650👀80742📊130

⚡ Powered by xgo.ing

If you know the team at NVIDIA that do the NVFP4 checkpoints, please intro me?

Suhail (@Suhail) — Tue, 25 Aug 2026 02:52:21 GMT

If you know the team at NVIDIA that do the NVFP4 checkpoints, please intro me?

💬5🔄0❤️10👀1842📊5

⚡ Powered by xgo.ing

Hi, does anyone know anyone on the DeepSeek team I can talk to? Interested in various models.

Suhail (@Suhail) — Mon, 24 Aug 2026 16:19:38 GMT

Hi, does anyone know anyone on the DeepSeek team I can talk to? Interested in various models.

💬7🔄0❤️26👀10376📊9

⚡ Powered by xgo.ing

AI that makes AI better (inference systems, training models, etc) is the only sensible benchmark to ...

Suhail (@Suhail) — Mon, 24 Aug 2026 00:57:07 GMT

AI that makes AI better (inference systems, training models, etc) is the only sensible benchmark to me moving forward.

All other work is downstream and possibly disrupted anyway.

💬24🔄4❤️105👀16730📊33

⚡ Powered by xgo.ing

Victory.

Suhail (@Suhail) — Sun, 23 Aug 2026 23:01:31 GMT

Victory.

💬2🔄1❤️14👀3304📊4

⚡ Powered by xgo.ing

This is my desktop now. I have 7 random text edits holding prompts for my agents for different situ...

Suhail (@Suhail) — Sun, 23 Aug 2026 22:58:38 GMT

This is my desktop now.

I have 7 random text edits holding prompts for my agents for different situations. Surely, this is not the future and a sign of how early it is.

💬161🔄16❤️403👀132663📊193

⚡ Powered by xgo.ing

We're back to this again

Suhail (@Suhail) — Sun, 23 Aug 2026 19:13:31 GMT

We're back to this again

💬15🔄1❤️98👀20333📊22

⚡ Powered by xgo.ing

Generally, I think this is quite a healthy thing overall.

Suhail (@Suhail) — Sun, 23 Aug 2026 18:03:56 GMT

Generally, I think this is quite a healthy thing overall.

💬1🔄1❤️13👀3158📊2

⚡ Powered by xgo.ing

New Chinese model subsidization at the inference layer is doing wonders for all the new agentic codi...

Suhail (@Suhail) — Sun, 23 Aug 2026 17:44:57 GMT

New Chinese model subsidization at the inference layer is doing wonders for all the new agentic coding agent products that compete with Anthropic/OpenAI.

💬32🔄10❤️133👀20332📊41

⚡ Powered by xgo.ing

Has saved me a lot of pain recently debugging issues: “Moving forward, when you encounter a bug or i...

Suhail (@Suhail) — Sun, 23 Aug 2026 12:51:27 GMT

Has saved me a lot of pain recently debugging issues: “Moving forward, when you encounter a bug or issue, search GitHub PRs, issues with another agent in parallel to see if you can find a reference solution.”

Which made me think, GitHub is the primary place agents are likely to already be communicating with one another.

💬20🔄6❤️144👀20083📊37

⚡ Powered by xgo.ing

16/ a bunch of people starting next week This one super annoying bug is killing us

Suhail (@Suhail) — Sun, 23 Aug 2026 01:55:24 GMT

16/ a bunch of people starting next week

This one super annoying bug is killing us

💬1🔄0❤️0👀3352📊1

⚡ Powered by xgo.ing

ok i shall do it myself

Suhail (@Suhail) — Sat, 22 Aug 2026 19:48:07 GMT

ok i shall do it myself

💬2🔄1❤️6👀6719📊3

⚡ Powered by xgo.ing

Has anyone gotten deepseek v4 flash 0731 to work with miles for training?

Suhail (@Suhail) — Sat, 22 Aug 2026 19:25:36 GMT

Has anyone gotten deepseek v4 flash 0731 to work with miles for training?

💬20🔄2❤️69👀21636📊24

⚡ Powered by xgo.ing

Cannot. Keep. Up. With. All. These. Models.

Suhail (@Suhail) — Fri, 21 Aug 2026 16:22:52 GMT

Cannot. Keep. Up. With. All. These. Models.

💬36🔄6❤️173👀22091📊46

⚡ Powered by xgo.ing

Free usage/inference for a month for DeepSeek V4 Flash 0731. Testing some things in our infra. DM. ...

Suhail (@Suhail) — Fri, 21 Aug 2026 01:32:20 GMT

Free usage/inference for a month for DeepSeek V4 Flash 0731. Testing some things in our infra.

DM. This tweet will be deleted.

💬9🔄1❤️27👀2982📊10

⚡ Powered by xgo.ing

Just make great things. Find your moat later.

Suhail (@Suhail) — Thu, 20 Aug 2026 23:34:59 GMT

Just make great things. Find your moat later.

Nikunj Kothari@nikunj

the models have no moat (OpenAI, Anthropic, XAI)

the IDEs have no moat (Cursor, Windsurf)

the harnesses have no moat (Cognition, Factory, LangChain)

the app builders have no moat (Replit, Lovable, Bolt)

the wrappers have no moat (Harvey, Abridge, OpenEvidence)

the inference providers have no moat (Together, Fireworks, Groq)

the voice layer has no moat (Sierra, Decagon, ElevenLabs)

the data labeling companies have no moat (Scale, Surge, Mercor)

the AI infrastructure has no moat (Baseten, Modal, Railway)

the neoclouds have no moat (CoreWeave, Lambda, Crusoe)

the generative media companies have no moat (Runway, Higgsfield, Suno)

apparently nobody in AI has a moat

except the venture firm ☠️

🔗 View Quoted Tweet

💬30🔄24❤️406👀82806📊64

⚡ Powered by xgo.ing

Are you using DeepSeek V4 Flash in production?

Suhail (@Suhail) — Thu, 20 Aug 2026 21:25:37 GMT

Are you using DeepSeek V4 Flash in production?

💬84🔄5❤️134👀37426📊85

⚡ Powered by xgo.ing

You need to have more than 10K WAU. They don’t need to be paying.

Suhail (@Suhail) — Thu, 20 Aug 2026 19:40:44 GMT

You need to have more than 10K WAU. They don’t need to be paying.

💬0🔄1❤️10👀3552📊1

⚡ Powered by xgo.ing

If you have a agentic coding app that has users but never quite reached scale due to competition, I’...

Suhail (@Suhail) — Thu, 20 Aug 2026 15:44:26 GMT

If you have a agentic coding app that has users but never quite reached scale due to competition, I’d like to potentially buy it. 

Pls dm.

💬37🔄10❤️222👀45110📊60

⚡ Powered by xgo.ing

Pizzamaxxing for my birthday this week.

Suhail (@Suhail) — Mon, 17 Aug 2026 18:48:17 GMT

Pizzamaxxing for my birthday this week.

💬13🔄1❤️75👀15097📊17

⚡ Powered by xgo.ing

❤️

Suhail (@Suhail) — Mon, 17 Aug 2026 01:34:43 GMT

❤️

💬9🔄0❤️39👀16483📊12

⚡ Powered by xgo.ing

@reinpk All this unbundling behavior—in the limit—simply causes an exploration of differentiation su...

Suhail (@Suhail) — Sun, 16 Aug 2026 18:37:41 GMT

@reinpk All this unbundling behavior—in the limit—simply causes an exploration of differentiation such that we are back to bundling.

I would know, Segment was one of the earliest companies to do this and they did so with Mixpanel!

💬2🔄1❤️5👀4532📊3

⚡ Powered by xgo.ing

@reinpk I blame you for all this. Segment was the OG gateway.

Suhail (@Suhail) — Sun, 16 Aug 2026 18:23:20 GMT

@reinpk I blame you for all this. Segment was the OG gateway.

💬3🔄0❤️7👀5952📊3

⚡ Powered by xgo.ing

This AI gateway thing is getting out of control. The advent of vibe coding creates -> 50 compan...

Suhail (@Suhail) — Sun, 16 Aug 2026 17:18:55 GMT

This AI gateway thing is getting out of control. 

The advent of vibe coding creates -> 50 companies -> people make gateway company (LLM, Sandbox, etc.) -> 50+ gateway companies -> make a gateway gateway company -> ...?

💬50🔄9❤️154👀41829📊63

⚡ Powered by xgo.ing

In NYC next week - best new bar/pizza in the last 24 mo?

Suhail (@Suhail) — Sat, 15 Aug 2026 22:22:19 GMT

In NYC next week - best new bar/pizza in the last 24 mo?

💬58🔄2❤️84👀130794📊74

⚡ Powered by xgo.ing

15/ the lowest it'll ever be

Suhail (@Suhail) — Sat, 15 Aug 2026 15:04:15 GMT

15/ the lowest it'll ever be

💬1🔄0❤️2👀4060📊1

⚡ Powered by xgo.ing

Are there competitors to OpenRouter?

Suhail (@Suhail) — Fri, 14 Aug 2026 21:46:34 GMT

Are there competitors to OpenRouter?

💬262🔄6❤️586👀199251📊306

⚡ Powered by xgo.ing

From 93K to a beautiful 3.1K LoC ❤️

Suhail (@Suhail) — Fri, 14 Aug 2026 20:34:42 GMT

From 93K to a beautiful 3.1K LoC ❤️

💬20🔄0❤️128👀54941📊29

⚡ Powered by xgo.ing

Looking for more great model optimizers / inference performance people for a new company I am buildi...

Suhail (@Suhail) — Thu, 13 Aug 2026 18:28:28 GMT

Looking for more great model optimizers / inference performance people for a new company I am building.

You’ll get to see a startup go from 0 to 100 very fast. We’ve locked down our compute.

In-person in SF. DM me.

💬10🔄4❤️107👀28290📊23